Surgeons' Acceptance of Surgical Site Infection Risk Adjustment Models

Background: To account for the wide range of reported surgical site infection (SSI) rates, current CDC/NHSN risk stratification models were designed to predict the expected risk of SSI in different groups of patients.  Surgeon acceptance of comparative data and risk adjustment is critical to drive quality improvement. 

Objective: The purpose of this focus group was to assess surgeons’ acceptance of current risk stratification models and to determine what factors surgeons felt are important for future models of SSI risk stratification.

Methods: Six academic surgeons with a research interest in SSIs, along with a facilitator, participated in a face-to-face focus group discussion held adjunct to the 5th Annual Academic Surgical Congress in San Antonio, TX.  The surgeons represented multiple health system types, including academic (4), private (2), safety net (2), and veterans administration (1).  Data were collected through audio recordings, transcribed notes, and observations by a qualitative researcher during the focus group session. Content analysis was conducted using an inductive approach with initial topics mentioned by participants identified through an open, heuristic coding process.  Topics identified as duplicative were combined into a single occurrence. A topic was identified as a theme if at least 3 participants mentioned or agreed with a topic. Data were reviewed to the saturation point and discussed with the facilitator to ensure the most comprehensive identification of patterns. 

Results: Multiple themes relevant to SSI risk stratification were identified and combined into 5 core concepts: (1) current risk adjustment models are inadequate; (2) surgeons acknowledged that they track an excessive number of variables that they feel contribute to SSI but emphasized the importance of focusing on factors that greatly affect the rate of SSI; (3) data are needed to determine which factors are most important and significant to the development of SSI, and risk stratification metrics must reflect this new data; (4) risks of SSI associated with emergency surgery should be considered separately from those of elective surgery; and (5) heterogeneous documentation (i.e., private vs public hospitals) affects the ability to both track and generalize the effects of potential risk factors on development of SSI.

Conclusions: Participating surgeons, representing multiple health system types, felt that the current models for SSI risk assessment are inadequate. While they believe that there are numerous factors that are likely important to accurately adjust risk, they agreed that many presumed risk factors may prove unimportant. A more refined risk adjustment model would likely improve surgeons’ acceptance of benchmarks and comparative data, and it may also drive quality improvement. These data suggest that further efforts at risk factor assessment are worthwhile.